Comparative study of hypertension, diabetes, dementia and smoking in military veterans and non-veterans: a quantitative study using primary healthcare data

Abstract

Introduction

Primary healthcare (PHC) patient medical records contain Systematised Nomenclature of Medicine-Clinical Terms (SNOMED-CT) that include information regarding diagnosis, demographics and veterans’ status. This study intended to identify, analyse and compare the prevalence of type 2 diabetes, hypertension, dementia and smoking tobacco in veterans and non-veterans, including stratification by age and gender.

Methods

The authors partnered with 13 PHC practices with a population of 137 410 patients. Staff extracted matched veteran and non-veteran SNOMED-CT data from patient medical records; then sent the authors anonymised data in an amalgamated format between October 2023 and January 2024. Patients were from a local community and therefore social and environmental factors would be similar. Submitted information was inputted into an SPSS database 28 for analysis which included descriptive and inferential statistics to indicate statistical significance.

Results

In total, 5458 PHC electronic records were examined comprising 2729 veterans and 2729 demographically matched for age and gender non-veterans. Each group contained 86.4% (N=2359) men and 13.6% (N=370) women. The mean age was 63.8 years (SD 17.7). Rates of hypertension were 20.9% in veterans compared with 17.6% in non-veterans (p=0.002). Type 2 diabetes mellitus was 8.3% in veterans compared with 6.4% in non-veterans (p=0.007). Dementia was 2.1% of veterans compared with 2.5% of non-veterans (p=0.32). Smoking was 11.8% of veterans compared with 10.6% of non-veterans (p=0.16).

Conclusion

These results reveal that veterans were statistically more likely to be diagnosed with hypertension and diabetes. This study should assist in a better understanding of the healthcare needs of the veteran population to potentially inform better patient-centred care. However, the effectiveness of using PHC patient medical records requires increased efforts to improve data quality which needs improved PHC staff knowledge, consistency in SNOMED-CT coding, better veteran medical e-record registration and coding and better data transmission between the Defence Medical Services and PHC.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Using primary healthcare (PHC) data can provide an original insight into veteran morbidity. This comparison study is the first of its kind to present UK veterans and non-veterans morbidity by using PHC medical records.

WHAT THIS STUDY ADDS

• Hypertension was 20.9% in veterans compared with 17.6% in non-veterans, and diabetes was 8.3% in veterans compared with 6.4% in non-veterans. Both were statistically significant.

• Comorbid hypertension and diabetes were presented in 27% of male and 23% of female veterans compared with 23% of male and 19% of non-veteran females.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• Effective using of PHC data requires knowledgeable PHC staff, consistent SNOMED (Systematised Nomenclature of Medicine-Clinical Terms) veteran coding and improved transmission of data between the Defence Medical Services and PHC.

• Statutory and non-statutory initiatives can tailor interventions to healthier lifestyles, including gym passes and/or connection to health promotion activities with professional sports clubs.

Introduction

There are estimated to be 1.85 million British Armed Forces Veterans living in England and Wales, which is approximately 3.8% of the total population.¹ The UK armed forces veteran population differs substantially from non-veterans in relation to gender and age distributions, with veterans being predominantly men (86%) and older, with those aged 80 years and over (34.8%) in comparison with the non-veterans (5.2%).¹ The Armed Forces Covenant was introduced to ensure there is no disadvantage to serving personnel, veterans and their families. Veterans should receive priority healthcare treatment and access to veteran-specific services when it relates to military service.²

Primary healthcare (PHC) is the portal to the NHS providing 300 000 million consultations each year.³ PHC patient electronic medical records use Systematised Nomenclature of Medicine-Clinical Terms (SNOMED-CT)⁴ that contain clinical information related to patient diagnosis, demographics (age, gender, ethnicity), diagnosis, laboratory tests and medications. SNOMED-CT codes extend to military veterans, and provide a consistent vocabulary for recording patient clinical information. SNOMED-CT codes inform the Quality and Outcomes Framework (QOF) for General Practitioner (GP) practices in England, Wales and Northern Ireland which is a voluntary annual reward and incentive programme,⁵ and provides information to indicate the levels of clinical morbidity. Therefore a SNOMED-CT search should reveal the healthcare status of the veteran population, although evaluations identified that only 9% were correctly coded.⁶ An initiative to improve registration resulted in 27% being coded,⁷ and included accessing PHC records to identify the levels of common mental health (MH) disorders in military veterans.⁸ However, it was unclear whether the prevalence was higher than in the non-veteran population.

Veterans overall seem not to differ from the non-veteran population in terms of the wider determinants of health.¹ However, male service personnel may feel unable to ask for help due to feelings of shame, stigma and fear for their career,⁹ with military culture and identity cited as reasons why veterans view any form of ill health as a sign of weakness.¹⁰ Therefore, a sense of pride may stop veterans from seeking help for any form of ill health.¹¹ Female veterans have also adopted the same attitude of not wanting to appear weak.¹² The Royal College of General Practitioners Veteran Friendly practice initiative intends to improve the recording of PHC support to veterans although the number of correctly registered veterans remains unknown.¹³

The aim was to identify and compare the prevalence of type 2 diabetes (T2DM), hypertension, dementia and smoking tobacco in veterans and non-veterans including stratification by age and gender.

Method

There are many different SNOMED-CT diagnostic codes available, and it is at the discretion of the staff to choose the appropriate code. Measures were taken to ensure consistency, including visits to PHC partners to discuss the requirements and any potential challenges. In April 2023, online meetings were offered to the PHC practices, followed by a final meeting in May 2023. Then the PHCs formed collaborative partnerships in developing and confirming their medical record search criteria, which was bolstered by those PHCs working in the same primary care network. As the PHC registered patients were from a local community then social and environmental factors such as housing, finance and education would be similar which would assist in the reliability and validity of the results.

Being mindful of the considerable pressures on PHC, the search strategy was limited to a small number of disorders. Staff extracted matched veteran and non-veteran SNOMED-CT data from PHC patient medical records; then anonymised by removing the patient’s name and individual NHS identification number, before sending it to the authors in an amalgamated format. The veteran inclusion criteria were any regular or reservists (including National Service) who had served at least 1 day in the British Armed Forces. The searches were submitted to the authors’ between October 2023 and January 2024.

The authors partnered with 13 Cheshire and Lancashire (England) PHC practices with a population of 137 410 patients. The authors had worked with the majority of these practices in previous studies, and they were identified as having strong veteran registration and/or being a large PHC practice (>10 000 patients).⁵ The PHCs were financially remunerated for their time. Detail from the study by Office of National Statistics¹ indicated that the veteran population in North West England was 3.8%, and this figure was used to estimate the veteran patients expected within each collaborating PHC practice, where the mean was 52% that were correctly coded, ranging from 28% to 92%. Submitted matched information was inputted into an SPSS database 28 for analysis which included descriptive statistics of frequencies and percentages, and inferential χ² test statistics to indicate statistically significant differences in the prevalence of disorders between the veteran and non-veteran populations.

Results

In total, 5458 PHC electronic records were examined comprising 2729 veterans and 2729 demographically matched for age and gender non-veterans. Each group contained 86.4% (N=2359) men and 13.6% (N=370) women. The mean age was 63.8 years (range of 18–100), medium 64, mode 84 and SD was 17.7 (figure 1).

Figure 1. Matched sample age of veterans and non-veterans.

Morbidity rates

Rates of hypertension were 20.9% (N=570) in veterans compared with 17.6% (N=480) in non-veterans. The difference was statistically significant at X²(1, N=5458)=9.5 p=0.002. Rates of T2DM were 8.3% (N=226) in veterans compared with 6.4% (N=174) in non-veterans. The difference was statistically significant at X² (1, N=5458)=7 p=0.007. Rates of dementia were 2.1% (N=58) compared with 2.5% (N=69) in non-veterans. The difference was not statistically significant at X² (1, N=5458)=0.975 p=0.32. Rates of smoking were 11.8% (N=321) in veterans compared with 10.6% (N=289) in non-veterans. The difference was not statistically significant at X² (1, N=5458)=2 p=0.16 (figure 2).

Figure 2. Physical health and smoking across veterans and non-veterans.

Gender

Among male veterans, 21.1% (N=498) were diagnosed with hypertension compared with 18% (N=424) of male non-veterans. A significant association at X² (1, N=4718)=7.38, p=007. Female veterans were 19.5% (N=72) diagnosed with hypertension compared with 15.1% (N=56) of female non-veterans but this was not significant X² (1, N=740)=2.42, p=0.120. Among male veterans, 8.6% (N=203) were diagnosed with T2DM compared with 6.5% (N=152) of male non-veterans which was statistically significant at X² (1, N=4718)=8 p=0.005. Amongst female veterans 6.2% (N=23) were diagnosed with T2DM compared with 5.4% (N=20) of female non-veterans X² (1, N=740)=0.24 p=0.637. When combining the diagnosis of hypertension and T2DM, then 27% of male veterans (N=629) and 23% (N=84) of female veterans had both compared with 23% (N=530) of non-veteran males and 19% (N=69) non-veteran females. The relationship between gender and the presence of T2DM and hypertension in veterans was not significant X² (1, N=2729)=2.60, p=0.107). Dementia was present in 2.2% (N=53) of male veterans compared with 2.5% (N=59) of non-veteran males. Dementia was present in 1.4% (N=5) of female veterans, and 2.7% (N=10) of non-veteran females. The relationship between dementia and gender in veterans was not significant X² (1, N=2729)=1.23, p=0.267). There were 12% of male veterans who smoked tobacco (N=283), compared with 10.6%% (N=251) of non-veteran males which was not significant X² (1, N=4718)=2.2, p=0.14. Of females veterans, 10.5% (N=39) were smokers compared with 10.3% (N=38) of female non-veterans X² (1, N=740)=0.11, p=1) (figure 3). Diagnostic levels between veterans and non-veterans by gender is shown in figure 4.

Figure 3. Health diagnosis and smoking by gender.

Figure 4. Health diagnosis and smoking in veteran and non-veterans by gender.

Age

In veterans, hypertension was most prevalent in the 88–96 age group (43.4%, N=105), followed by the 78–87 age group (29.4%, N=145). Differences across age groups and hypertension were significant X² (8, N=2729)=228.8, p<0.01). T2DM was most prevalent in the 58–67 (10.7%, N=59) and 88–96 (10.7%, N=26) age groups. Differences across age groups and T2DM were significant X² (8, N=2729)=38.01, p<0.01). Dementia was most prevalent in the 88–96 age group (12%, N=29) with differences across age groups being significant X² (8, N=2729) =161.6, p<0.01). Smoking was most prevalent in the 18–27 age group (17%, N=8). Differences across age groups and smoking were significant X² (8, N=2729)=22.3, p<0.01) (figure 5).

Figure 5. Percentage of health diagnosis and smoking in veterans by age groups.

Smoking, hypertension and diabetes

There were significant differences between veterans and non-veterans who were coded for smoking and hypertension (F(1, 5458)=5.490, p=0.002), with non-veterans at (12.1%, N=231) and veterans (9.5%, N=268). Also, non-veterans (14.3%, N=264) were more likely than veterans (10.2%, N=299) to be coded for smoking and T2DM but this difference was not significant (F(1, 5458)=4.082, p=0.084) (table 1).

Table 1. Smoking association with hypertension and type 2 diabetes in veterans and non-veterans.

Smoking and hypertension
Veteran or non-veteran			Hypertension				Total
			No		Yes
			N	%	N	%	N	%
Veteran	Smoker	No	1891	87.6	516	90.5	2407	88.2
		Yes	268	12.4	54	9.5	322	11.8
	Total		2159	100.0	570	100.0	2729	100.0
Non-veteran	Smoker	No	2018	89.7	422	87.9	2440	89.4
		Yes	231	10.3	58	12.1	289	10.6
	Total		2249	100.0	480	100.0	2729	100.0

Smoking and type 2 diabetes mellitus
Veteran or non-veteran			Diabetes				Total
			No		Yes
			N	%	N	%	N	%
Veteran	Smoker	No	2204	88.1	203	89.8	2407	88.2
		Yes	299	11.9	23	10.2	322	11.8
	Total		2503	100.0	226	100.0	2729	100.0
Non-veteran	Smoker	No	2291	89.7	149	85.6	2440	89.4
		Yes	264	10.3	25	14.3	289	10.6
	Total		2555	100.0	174	100.0	2729	100.0

Discussion

Hypertension

It is estimated that 11.8 million adults aged 16 years or older in England had hypertension, which is mainly detected and managed in general practice.¹⁴ Primary hypertension occurs in about 90% of people and has no identifiable cause, leaving about 10% of people with an underlying cause, such as renal, endocrine or vascular disorder. Generally, there are no symptoms, but early diagnosis and effective management can curtail the long-term effects of hypertension including the most common risk factor for cardiovascular disease.¹⁵ The prevalence of hypertension increases with age and is more common in deprived areas.¹⁴

National Institute for Health and Care Excellence (2023) hypertension guidelines¹⁶ have a diagnosis aligned to a clinic-recorded BP of 140/90 mm Hg or higher, and either a subsequent ambulatory BP monitoring daytime average or home BP monitoring average of 135/85 mm Hg or higher. Other organisations have different thresholds such as the American College of Cardiology/American Heart Association of at least 130/80 mm Hg¹⁷ and US primary care≥130 mm Hg or DBP≥90 mm Hg.¹⁸ However, lowering the BP diagnostic and treatment cut-offs significantly increases the burden on healthcare systems¹⁵ and may in part explain why our partnering Cheshire practices use 150/90 mm Hg. This clearly impacts reporting incidence levels.

USA research reveals combat exposure and/or combat injury may increase the risk of hypertension through changes in inflammatory responses, psychological stress and health behaviours.¹⁹ Howard et al²⁰ analysis of Department of Defence Trauma Registry data indicated that post-traumatic stress disorder (PTSD), being overweight and obesity were associated with higher odds of hypertension, and that combat exposure increases hypertension risk and that combat injury exacerbates this risk. In this study, rates of hypertension were 20.9% in veterans compared with 17.6% in non-veterans. This was statistically significant; equating to 480 700 UK veterans with hypertension. Male veterans at 21.1% were higher when compared with 18% of male non-veterans; and female veterans were 19.5% compared with 15.1% of female non-veterans. As expected, percentages rose incrementally with age.

Military lifestyle

Veterans are at high risk for uncontrolled BP and its complications,¹⁵ and there is a complex interplay of certain comorbid conditions such as T2DM that suggest the need for multifaceted interventions to prevent cardiovascular events.²¹ Military personnel experience significant occupational demands of high physical exercise regimens, often supplemented with heavy manual work for the combat and front-line career employment groups. The aligned calorific requirement to maintain this level of physical endeavour is excessive and service personnel habitually have a high calorific intake without fear of being overweight. That level of fitness is maintained and monitored with annual mandatory personal fitness tests, which when failed results in remedial training. However, on leaving the armed forces, for many that regime disintegrates, yet their calorific consumption may remain the same while they may progress into more sedentary employment. Combined with getting older, weight gain is inevitable unless preventative measures are introduced. Statutory and non-statutory initiatives can tailor a specific course of interventions to healthy lifestyles. A proactive measure is to provide veterans with gym passes and/or connection to health promotion activities with iconic organisations such as professional sports clubs.²² As 150 000 veterans access over 500 Armed Forces Breakfast Clubs, then aligned public health notices or initiatives could help. Unfortunately, poor diet is associated with poor finances, and working-age veterans need well-paid employment.

Type 2 diabetes mellitus

QOF (2024) data indicates that 7.5% of adults were recorded as having T2DM and the percentage was higher in deprived areas.²³ In this study, rates of T2DM were 8.3% in veterans compared with 6.4% in non-veterans. The difference is statistically significant. This equates to 190 900 UK veterans with T2DM. Among male veterans, 8.6% were diagnosed with T2DM compared with 6.5% of male non-veterans which was again statistically significant.

The prevalence of doctor-diagnosed T2DM increased with age, from 1% of adults aged under 35 to 16% of those aged 75 and over.²⁴ In our study, T2DM was most prevalent in the 58–67 and 88–96 years old veterans both at 10.7% each. T2DM in military veterans was higher in all age groups except for 38–47, and notably higher from the age of 48 years on onwards.

A retrospective cohort study of a large national sample in Scotland involved up to 37 years follow-up and matched for age, sex and area of residence. The results indicated that 7.2% of veterans were diagnosed with T2DM and were at slightly increased risk compared with non-veterans. The increased risk was confined to men, and to veterans born prior to 1960. There was a positive association between T2DM and PTSD, especially in the presence of comorbid mood disorder.²⁵

Stefanovics et al²⁶ completed a nationally representative sample of US veterans to assess PTSD and obesity prevalence, co-occurrence and relationships with mental and physical health measures. A total of 16.4% of veterans screened positive for current PTSD, 32.7% for obesity and 5.8% for co-occurring PTSD and obesity. Relative to obesity-only veterans, veterans with co-occurring PTSD and obesity had elevated likelihoods of mental and physical health concerns (most notably major depressive and generalised anxiety disorders), suicidality and migraine headaches, and higher body mass indices. Relative to veterans with PTSD alone, individuals with comorbid PTSD and obesity had elevated likelihoods of suicidal ideation, nicotine dependence, MH treatment, migraine headaches, diabetes, hypertension and insomnia. A significant minority of US veterans have co-occurring PTSD and obesity, which is associated with substantial mental and physical health burden, including elevated suicidality. Given our study reveals a higher prevalence of veterans with T2DM, then further UK research should explore the association with MH disorders.

Dementia

In the UK, there are an estimated 1 in 11 people over the age of 65 years who have dementia, which is the leading cause of death with a population estimate of 7.1%.²⁷ USA research highlighted an increased prevalence of dementia in veterans, particularly when associated with PTSD,²⁸ while UK research indicated that rates in the veteran population were no higher than non-veterans.²⁹ This study identified 2.1% of veterans recorded as having dementia compared with 2.5% of non-veterans; equating to 48 300 UK veterans living with dementia. As expected, dementia was most prevalent in the 88–97 years old group and above at 12%. However, the Census 2021 indicated that nearly one-third of veterans (31.8%) were aged 80 years or older,¹ so this figure remains lower than the authors expected to find.

Chung et al³⁰ conducted a UK longitudinal cohort analysis of linked electronic health records from 4.3 million people regarding dementia incidence, comorbidities, reasons for health‐care visits, mortality, causes of death and examined dementia patterns by relative deprivation in the UK. Reasons for hospitalisation and causes of death were compared in individuals with and without dementia. From 1998 to 2016, 145 319 (3.1%) of individuals were observed with incident dementia. Repeated hospitalisations among senior adults for infection, unknown morbidity and multiple PHC visits for chronic pain were observed prior to dementia diagnosis, and multiple long‐term conditions were present in half of the individuals at the time of diagnosis. Individuals living in high-deprivation areas had higher dementia incidence and high fatality. They concluded that there was considerable disparity of dementia that informs priorities of prevention and provision of patient care.

There are multiple reasons why elderly veterans’ PHC e-medical records were not correctly coded. These include elderly veterans in particular being unaware of the 1-day inclusion criteria^{7 8} and therefore ignorant of their veteran status and never declaring as such. The study’s PHC staff suggested that dementia diagnosis is often only confirmed after a long insidious presentation, often with family and other support being in place. Large numbers of vulnerable veterans may be residing in care/residential homes, and they may be reliant on others for communication with PHC. In addition to ensuring that this information can be used to provide the best clinical care; then it also has important ramifications for non-statutory and third sector organisations who have considerable resources to assist all of the armed forces community. The authors have previously highlighted that women on average live to an older age than men (UK: 83 compared with 79 years), and veterans’ spouses may be socially isolated and with financial difficulties.⁸

Smoking

Rates of smoking tobacco were higher in veterans (11.8%) than non-veterans (10.6%), being more so in males, and most prevalent in the 18–27 age group at 17%. This equates to 271 400 veterans who smoke tobacco. In smokers, the results indicated that the association between T2DM and hypertension was higher in non-veterans although the numbers were low. Smoking tobacco carries a high public health risk²⁵ and often commences or is maintained while serving in the armed forces, and efforts should be continued to address and offer early support.

Lessons learnt and future research

The effectiveness of exploiting PHC patient medical records requires increased efforts to improve the four areas that will lead to better quality data, that being better PHC staff knowledge of veterans, consistency in PHC SNOMED-CT coding, better veteran medical e-record registration and coding, with better transmission of data between the Defence Medical Services (DMS) and NHS services.

Military medical screening would have restricted enlistment with conditions including diabetes, and as the study does not reveal the causes then it is unclear if the medical conditions were aligned to military service. A study that assesses written PHC patient notes would prove enlightening to specifically identify what situational factors are impacting the veteran population. The study methodology can be extended to other physical and MH disorders.

Limitations included that classification as a UK ‘veteran’ requires an individual to serve for only 1 day, and the results include those unlikely to be affected by military service. Furthermore, the coding of a disorder can indicate a lifetime prevalence and at times it was difficult to differentiate between enduring or resolved conditions. Therefore, the results could be inflated with people who were successfully treated being categorised as active cases. Multiple SNOMED-CT codes for the same condition impact on consistency of reporting, and DMS patient records are not always added to PHC records. The study was in North West England which included variance across social scales of deprivation, but there are wider differences for veterans living in other areas and Celtic countries. PHCs received a financial incentive.

Conclusion

This comparison study is the first of its kind to present UK veteran and non-veterans morbidity by using PHC medical records, and provides information to inform health organisations and other relevant stakeholders to identify gaps in service provision. This should assist in understanding the healthcare needs of the veteran population, and potentially provide better patient-centred care within all healthcare sectors including NHS England’s Op RESTORE. The results should also be considered in developing veterans-specific educational syllabus and policy.

Data availability statement

Data are available upon reasonable request.